EARLY STAGES OP DEVELOPMENT OF THE MOUSE. 71 



Tarsius are concerned, all of which I have myself studied. I 

 believe Duval's hypothesis to be quite uuteuable, particularly 

 in the last-mentioned case, in which the epiblastic knob is 

 differentiated long before the hypoblast grows out to form 

 the umbilical vesicle. Erinaceus, it must be admitted, is an 

 exception, in so far as here the epiblastic knob does not begin 

 to form until some time after the hypoblast is completely 

 separated off; but we must at the same time remember that 

 the earliest stages in the development of this mammal are 

 quite unknown to us. 



We may now turn to a consideration of the far-reaching 

 speculations of Robinson (25) and Assheton (2). Of these 

 Assheton's is partly based on that of Robinson, while both 

 rest upon phenomena which their authors have, or supposed 

 themselves to have observed, in a single genus alone. 



I have already endeavoured to show that Robinson's 

 account of the early development of the mouse is devoid of 

 any serious basis in fact; and I need, therefore, now only 

 briefly discuss his extension of this erroneous description to 

 the rest of the Mammalia. 



Starting from a comparison of the blastocystic cavity with 

 the segmentation cavity, and of the embryonic knob with the 

 yolk-cells of the lower Vertebrata (an homology which might 

 be criticised on more grounds than one), he finally arrives at 

 the conclusion that the epiblast forms a small knob on the 

 surface of a hypoblastic vesicle, round which it is unable to 

 grow as it does in the Sauropsida and in Erinaceus ; and 

 seizing on the statement of van Beneden and Julin (since, 

 however, contradicted by Duval), that in the bat the hypo- 

 blast does not completely grow round at the anti-embryonic 

 pole of the blastocyst, he puts forward the hypothesis that it 

 is in reality the edges of the epiblast which are here free, 

 while the hypoblast forms a completely closed vesicle round 

 which the epiblast grows, and not conversely (fig. H), and 

 this hypothesis he proposes to extend to the rabbit, mole, 

 and other mammals. This strained interpretation has been 

 already contested by Assheton, and a glance at his figures. 



